Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Hypoxic pulmonary vasoconstriction (HPV) is a fundamental lung vasoregulatory mechanism, matching blood flow through individual acini with ventilation and thereby optimizing gas exchange (1). Cell(s) responsible for O₂ sensing, sensor mechanism(s), and the pathway(s) of signal transduction to the contractile apparatus of the precapillary vascular smooth-muscle cells do, however, remain enigmatic. There is recent evidence for an involvement of nitric oxide (NO) generation in HPV (2). Exhaled NO sharply decreases upon onset of hypoxia and returns to baseline upon reoxygenation, and these metabolic changes precede the changes in vasomotor tone. Moreover, the strength of HPV is markedly amplified upon inhibition of lung NO synthesis. These data suggest that tonic lung NO synthesis helps to establish normoxic pulmonary vasodilatation, presumably via stimulation of soluble guanylate cyclase with formation of cyclic guanosine monophosphate (cGMP). The lowering of baseline NO generation in response to alveolar hypoxia contributes to the vasoconstrictor response in the hypoxic region of the lung, but the remaining NO synthesis still attenuates the maximum strength of the vasoconstrictor response. However, inasmuch as inhibition of lung NO synthesis in the absence of hypoxia does not mimic HPV, additional signaling events are apparently involved in this regulatory mechanism.

Investigations from different laboratories including our own (5) suggested that a nicotinamide-adenindinucleotide (phosphate) (reduced form) (NAD[P]H) oxidase- dependent superoxide generation with subsequent formation of H₂O₂ may represent such a signaling cascade. However, there is controversy as to the mode of action of this cascade, and two different suggestions have been offered. First, Burke and Wolin (11, 12) suggested that guanylate cyclase is stimulated by H₂O₂ (in association with catalase, forming compound I), and that tonic generation of superoxide and subsequent H₂O₂ under normoxic conditions may be operative to affect baseline formation of the vasodilatory cGMP, as is similarly assumed for the NO-cGMP axis. Loss of such normoxic superoxide/H₂O₂/ cGMP generation will then again result in a vasoconstrictor response to alveolar hypoxia (13). Second, investigations of oxygen radical formation in isolated pulmonary arteries and smooth-muscle cells and inhibitor studies in intact lungs (5, 10) provided evidence that superoxide and H₂O₂ formation might not be decreased but rather increased in response to alveolar hypoxia, thereby contributing to the hypoxic vasoconstrictor response independent of guanylate cyclase stimulation and cGMP synthesis.

To elucidate these controversies, the present study addressed the role of guanylate cyclase in the regulation of HPV in more detail. Perfused rabbit lungs were used for provocation of repetitive hypoxic or pharmacologic (stable thromboxane analogue U46619) vasoconstrictor responses, and the impact of guanylate cyclase inhibitors as well as a phosphodiesterase (PDE) type V inhibitor for prolongation of cGMP half-life was investigated. Studies were performed in the presence of NO and under conditions of preblocked lung NO synthesis to discriminate between the NO-cGMP axis and cGMP formation via an alternate route. In essence, the findings support an important role of NO-dependent guanylate cyclase activity in attenuating the vasoconstrictor response to alveolar hypoxia in rabbit lungs, whereas no evidence was obtained for a role of NO-independent cGMP formation in this vasoregulatory mechanism. This contrasts with the vasoconstrictor response provoked by U46619, which was observed to be counterbalanced via lung guanylate cyclase activity even under conditions of preblocked NO synthesis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents

LY83583 was provided by Calbiochem (Bad Soden, Germany), and acetylsalicylic acid (ASA) was obtained from Bayer (Leverkusen, Germany). U46619 was from Paesel+Lorei (Frankfurt, Germany). Methylene blue and Zaprinast were purchased from Sigma (Deisenhofen, Germany). The perfusate was provided by Serag-Wiessner (Naila, Germany). All other biochemicals were obtained from Merck (Munich, Germany).

Lung Isolation, Perfusion, and Ventilation

The model of isolated perfused rabbit lungs has been described previously (14). Briefly, pathogen-free rabbits of either sex (body weight 2.2 to 3.2 kg) were deeply anesthetized and anticoagulated with heparin (1,000 U/kg body weight). The lungs were excised while being perfused with Krebs-Henseleit buffer through cannulae in the pulmonary artery and the left atrium. The buffer contained 125.0 mM NaCl, 4.3 mM KCl, 1.1 mM KH₂PO₄, 2.4 mM CaCl₂, 1.3 mM MgCl₂, and 275 mg glucose per 100 ml; NaHCO₃ was adjusted to result in a constant pH range of 7.37 to 7.40. After rinsing the lungs with at least 1 liter of buffer fluid for washout of blood, the perfusion circuit was closed for recirculation (total system volume 350 ml). Meanwhile, the flow was slowly increased from 20 to 150 ml/min, and left atrial pressure was set at 1.5 to 2.0 mm Hg to ensure zone III conditions throughout the lung at end expiration. The alternate use of two separate perfusion circuits allowed repeated exchange of buffer fluid. In parallel with the onset of artificial perfusion, ventilation was changed from room air to a mixture of 5.3% CO₂, 21.0% O₂, and the balance N₂ (tidal volume, 30 ml; frequency, 30 strokes/min). A positive end-expiratory pressure of 1 cm H₂O was chosen (0 referenced at the hilum). The isolated perfused lungs were placed in a temperature-equilibrated housing chamber, freely suspended from a force transducer for continuous monitoring of organ weight. The whole system (perfusate reservoirs, tubing, and housing chamber) was heated to 38.5°C. Pressures in the pulmonary artery, the left atrium, and the trachea were registered by means of small-diameter tubing threaded into the perfusion catheters and the trachea and connected to pressure transducers. Lungs included in the study were those that (1) had a homogeneous white appearance with no signs of hemostasis, edema, or atelectasis; (2) revealed constant mean pulmonary artery pressure () and peak ventilation pressure in the normal range; and (3) were isogravimetric during the initial steady-state period of at least 20 min.

Hypoxic Maneuvers and Pharmacologic Challenges

The technique of successive hypoxic maneuvers in buffer-perfused rabbit lungs has been described previously (14). Briefly, a gas mixing chamber (KM 60-3/6MESO; Witt, Witten, Germany) was used for step changes in the ventilator O₂ content {21% vol/ vol [alveolar oxygen partial pressure (PA_O2) ~ 160 mm Hg, baseline conditions] to 3% vol/vol [PA_O2 ~ 23 mm Hg, hypoxic conditions]}. A 5.3% vol/vol quantity of CO₂ was used throughout, and the percentage of N₂ was balanced accordingly. Buffer returning from the perfusate reservoir to the lungs passed through a membrane oxygenator (M8Exp; Jostra, Hirrlingen, Germany). By this device the partial pressure was set at ~ 40 mm Hg for both CO₂ and O₂ in the postoxygenator buffer fluid entering the pulmonary artery. Sequential hypoxic maneuvers of 10 min duration, interrupted by 15-min periods of normoxia, were performed. The effects of the various pharmacologic agents on pressure responses provoked by alveolar hypoxia (3% O₂) were determined within such a sequence of repetitive hypoxic maneuvers. Each agent was added to the buffer fluid 5 min before a hypoxic challenge, commencing the addition after accomplishing the second hypoxic maneuver. Cumulative dose-effect curves were established. For comparison, the influence of these agents on U46619-elicited pressor responses was tested. In these experiments, a mode of repetitive bolus applications of the stable thromboxane analogue was used (addition to the perfusate at 0.5 nM every 25 min) as described previously (5, 10). For calculation of relative values, the response to the second vasoconstrictor provocation in a sequence of challenges was set at 100% (= reference response) for each lung preparation. The strengths of the following vasoconstrictor responses were related to this reference response. Control experiments were performed with use of the vehicle only. In the experiments with preblocked NO synthesis, 400 µM N^G-monomethyl-L-arginine (L-NMMA) and 1 mM ASA (to avoid any confounding influence of lung prostanoid synthesis) were added to the perfusate from the beginning of the experiments. Lung weight was continuously monitored, but the total weight gain ranged < 3 g in all experiments.

Addition of NO to the Perfusate

In the experiments with NO addition, the membrane oxygenator was used for equilibration of the buffer fluid with the gas (15). NO, CO₂, O₂, and N₂ were premixed in a gas mixing chamber (KM60-4 ME; Witt) with PO₂ and pCO₂ set at ~ 40 mm Hg. The NO concentration used to supply the membrane oxygenator was determined in each experiment using a chemiluminescence NO analyzer (Sievers 280 NOA; Sievers Instruments, Boulder, CO). High gas flow was used (gas to perfusate flow ~ 10:1) for full equilibration of partial pressures, targeting a perfusate NO concentration of 82 nM. The response to perfusate NO admixture was tested by the following mode: 10-min periods with hypoxic ventilation were performed, alternating with 15-min periods of normoxic ventilation. At 5 min before the third hypoxic challenge, equilibration of the perfusate with NO was started. The strength of the third hypoxic response was related to the second (= reference) response. Thereafter the lung was flushed with 1 liter of fresh perfusate and the same mode for assessment of pre- and post-NO vasoconstriction was performed for U46619-bolus applications (every 25 min, 0.5 nM) instead of hypoxic challenges. Randomization was undertaken, starting half of the experiments with hypoxic and half of the experiments with U46619 challenges. Controls received no NO.

Statistics

For comparison of statistical differences, analysis of variance with the Student-Newman-Keuls post hoc test was performed. Statistical significance was assumed when P ranged  0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Under baseline conditions, values were 5.6 ± 0.2 mm Hg (mean ± standard error of the mean [SEM], n = 32). A 3% hypoxic challenge (PA_O2 ~ 23 mm Hg) consistently provoked a rapid increase in Ppa, with pressure elevations of 2.8 ± 0.2 mm Hg (mean ± SEM, n = 20). Repetitive hypoxic challenges resulted in well-reproducible pressure elevations within the same lung, as previously described (14). Prior blockage of NO synthesis and cyclooxygenase by L-NMMA (400 µM) and ASA (1 mM) only marginally affected normoxic vascular tone, but increased hypoxic pressor responses throughout the experiments: pressure elevation in response to hypoxia was 6.9 ± 0.7 mm Hg (mean ± SEM, n = 12) under these conditions. Similarly, repetitive bolus application of U46619 provoked well-reproducible vasoconstrictor responses under baseline conditions (5.0 ± 0.6 mm Hg; mean ± SEM, n = 12), again enhanced in the presence of L-NMMA and ASA (10.9 ± 1.7 mm Hg, mean ± SEM, n = 12).

Addition of the guanylate cyclase inhibitor LY83583 slightly increased normoxic vascular tone (Table 1) and amplified HPV in a dose dependent manner (Figure 1). The U46619-elicited pressor response was enhanced in a corresponding manner. Admixture of LY83583 under conditions of preblocked NO and prostanoid synthesis, in contrast, did not further increase the pressor responses to hypoxia but again amplified vasoconstrictions evoked by U46619 (Figure 1). Corresponding results were obtained with the second guanylate cyclase inhibitor, methylene blue (Table 1, Figure 2). No differences were found between experiments performed with addition of L-NMMA alone or L-NMMA combined with ASA (data not shown).

View larger version (36K): [in this window] [in a new window]   Figure 1.   Influence of the guanylate cyclase inhibitor LY83583 on the strength of hypoxia- and U46619-induced vasoconstrictions during intact (left) and blocked (right) NO synthesis. Top shows relative values, referenced to the second vasoconstrictor response (= 100%). For depiction of absolute values (bottom), the control pulmonary arterial pressure (Ppa) response and the response in the presence of the highest inhibitor dose are displayed for each individual experiment. Change in Ppa (PAP) = strength of hypoxia- or U46619-elicited increase in Ppa. Cumulative concentrations are given. Relative values are means ± SEM, n = 4 lungs each. In controls, vehicle only was applied. *Statistically significant differences of the LY83583-treated groups when compared with their corresponding controls. +Statistically significant differences between the impact of LY83583 on hypoxia- versus U46619-elicited vasoconstriction.

View larger version (33K): [in this window] [in a new window]   Figure 2.   Influence of the guanylate cyclase inhibitor methylene blue on the strength of hypoxia- and U46619-induced vasoconstrictions during intact (left) and blocked (right) NO synthesis. Top shows relative values, referenced to the second vasoconstrictor response (= 100%). For depiction of absolute values (bottom), the control Ppa response and the response in the presence of the highest inhibitor dose are displayed for each individual experiment. Change in Ppa (PAP) = strength of hypoxia- or U46619-elicited increase in Ppa. Cumulative concentrations are given. Relative values are means ± SEM, n = 4 lungs each. In controls, vehicle only was applied. *Statistically significant differences of the methylene blue-treated groups when compared with their corresponding controls. +Statistically significant differences between the impact of methylene blue on hypoxia- versus U46619-elicited vasoconstriction.

The PDE V inhibitor Zaprinast (1 to 10 µM) resulted in a marginal decrease of normoxic vascular tone (Table 1). Using this dosage range, the hypoxia-elicited vasoconstrictor responses were significantly more inhibited than those in response to U46619 (Figure 3). Under conditions of preblocked NO and prostanoid synthesis this difference disappeared, and moderate inhibition was noted for both hypoxia- and U46619-provoked pressure elevation (Figure 3). Again, no differences were observed in experiments performed with L-NMMA alone or L-NMMA in combination with ASA (data not shown).

View larger version (31K): [in this window] [in a new window]   Figure 3.   Effects of the guanylate cyclase inhibitor Zaprinast on the strength of hypoxia- and U46619-induced vasoconstrictions during intact (left) and blocked (right) NO synthesis. Top shows relative values, referenced to the second vasoconstrictor response (= 100%). For depiction of absolute values (bottom), the control Ppa response and the response in the presence of the highest inhibitor dose are displayed for each individual experiment. Change in Ppa (PAP) = strength of hypoxia- or U46619-elicited increase in Ppa. Cumulative concentrations are given. Relative values are means ± SEM, n = 4 lungs each. In controls, vehicle only was applied. *Statistically significant differences of the Zaprinast-treated groups when compared with their corresponding controls. +Statistically significant differences between the impact of Zaprinast on hypoxia- versus U46619-elicited vasoconstriction.

Equilibration of gaseous NO with the perfusate (82 nM) reduced HPV to 17.7 ± 3.6% (mean ± SEM, n = 4) of the reference hypoxic challenge (Figure 4). The U46619-elicited pressor response was less prominently attenuated by this approach. Baseline vascular tone only marginally decreased in response to NO.

View larger version (12K): [in this window] [in a new window]   Figure 4.   Influence of perfusate equilibration with NO (82 nM) on hypoxia- and U46619-induced vasoconstriction. Values are compared with time-matched control experiments. Values are means ± SEM; n = number of lungs, n = 4 experiments each. *Statistically significant differences of the NO-treated groups when compared with their corresponding controls. +Statistically significant differences between the impact of NO on hypoxia- versus U46619-elicited vasoconstriction. Change in Ppa = PAP.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study focused on the role of NO-dependent versus NO-independent guanylate cyclase activity in oxygen-responsive pulmonary vasoregulation in rabbit lungs. The current findings are consistent with an involvement of the NO-cGMP axis in HPV, whereas no evidence was obtained for a role of NO-independent guanylate cyclase activity in the vasomotor changes in response to hypoxia. In contrast, both NO-dependent and NO-independent cGMP formation is suggested to attenuate the vasoconstrictor response to the stable thromboxane analogue U46619. These conclusions are based on the following observations:

First, baseline Ppa assessed under normoxic conditions, moderately increased upon inhibition of NO synthesis and upon blockage of lung guanylate cyclase activity with both methylene blue and LY83583. Corresponding activity was previously noted for methylene blue in intact-chest cats (16) and isolated rat pulmonary arteries (17). A large increase in normoxic vascular tone by LY83583 was described for small pulmonary arteries of the calf (18). The difference between the latter and the current investigation may be explained by the use of U46619 in the isolated pulmonary arteries to precontract these vessels. It is in line with the present study that under such conditions (presence of U46619) there is marked enhancement of vasoconstriction by U46619.

Second, guanylate cyclase inhibition by the two independent inhibitors methylene blue and LY83583 (16, 17, 19) markedly amplified the vasoconstrictor response to both hyp-oxia and U46619 to a similar extent. These agents were previously shown to inhibit cGMP synthesis in various systems including perfused lungs (17, 19, 20). Corresponding efficacy for enhancement of the pressor response to alveolar hypoxia was previously demonstrated for inhibition of NO synthesis in the current model of perfused rabbit lungs (2). Augmentation of the strength of HPV was similarly noted in methylene blue-treated isolated rat lungs (23).

Third, in the presence of 400 µM L-NMMA, previously shown to result in virtually complete blockage of rabbit lung NO synthesis (15), no additional impact of simultaneous guanylate cyclase inhibition on the hypoxia-elicited vasoconstrictor response was noted. This finding was true for both methylene blue and LY83583, but sharply contrasted to the thromboxane-provoked vasoconstriction: even under conditions of preblocked NO synthesis, the pressor responses to U46619 were significantly enhanced upon simultaneous guanylate cyclase inhibition by methylene blue or LY83583. These data strongly support the notion that NO-independent guanylate cyclase activity and cGMP formation are operative in attenuating the vasoconstrictor response to the pharmacologic agent but not the vasoconstrictor response to alveolar hypoxia. Our investigation is, however, in contrast to the finding of Omar and Wolin (18) that LY83583 mimics hypoxia in small pulmonary artery rings of the calf. Normoxic vascular tone was prominently increased and hypoxia-induced vasoconstriction was inhibited in that study by LY83583. The discrepancy with our investigation may again be related to the fact that U46619 was used to precontract the isolated vessel, as the present study clearly demonstrated that U46619-induced vasoconstriction is amplified by guanylate cyclase inhibitors. Thus, there is an overlap of U46619- and hypoxia-induced vasoconstriction in the study of Omar and Wolin (18) that makes interpretation of hypoxia effects difficult. Moreover, isolated vessels may react in a different way than isolated lungs, where natural cell-cell contacts are maintained. This argument is supported by the investigation of Mazmanian and colleagues (23), who found similar effects on HPV for the guanylate cyclase inhibitor methylene blue in isolated rat lungs as did the present study. Another difference in our investigation is the use of 0% oxygen for hypoxia in the studies of Omar and Wolin (18). In pulmonary arteries of the rat it has been shown that methylene blue inhibited contraction to 0% O₂ but increased the response to milder hypoxia (17). Further, our findings are supported by the corresponding effects of two distinct guanylate cyclase inhibitors in one experimental setup.

Fourth, the selective PDE V inhibitor Zaprinast (24), capable of prolonging the half-life of cGMP independent of the underlying route of guanylate cyclase activation, reduced the vasoconstrictor response to alveolar hypoxia significantly more prominently than that to U46619, and this significant difference was lost under conditions of preblocked lung NO synthesis. This observation again supports the higher significance of the NO-cGMP axis in the vasomotor changes provoked by hypoxia as compared with thromboxane-elicited vasoconstriction. Attenuation of the HPV by Zaprinast was previously also reported for rat lungs in one study (25) but not found in another study (26). The stable cGMP analogue 8-bromo-cGMP is known to decrease the strength of HPV in isolated rat lungs (20).

Fifth, and finally, in line with the suggestion for a more prominent role of NO-triggered guanylate cyclase activity in hypoxia- as compared with U46619-elicited vasomotion, the NO enrichment of the perfusate returning to the lung more potently inhibited the hypoxic vasoconstriction as compared with the pharmacologic vasoconstriction.

Changes in lung NO release upon step-change from normoxia to hypoxia have been directly demonstrated (2, 3, 27), whereas the evidence for a role of the NAD(P)H oxidase-superoxide-H₂O₂ axis in hypoxic pulmonary vasoconstriction is largely based on studies with various inhibitors (5, 7, 9, 10, 13). Even if a recent study denies a contribution of the gp91phox subunit of the phagocyte-type oxidase (28), a role of other NAD(P)H oxidase components or a NAD(P)H oxidase similar but not identical to that of phagocytes cannot be excluded. The currently reported lack of evidence for a NO-independent guanylate cyclase activation in HPV does not exclude any role of activated oxygen species in vasomotor changes to hypoxia, but it questions the hypothesis that H₂O₂ might be operative via stimulating guanylate cyclase to elicit normoxic vasodilation in a NO-independent fashion, the loss of which is one important component of the HPV (7, 8, 11). In contrast, the present observations are compatible with the recent suggestion that the superoxide/H₂O₂ formation might rather be increased than decreased in hypoxia (6, 29), favoring a mode of action of this system fully independent of H₂O₂-elicited guanylate cyclase activation.

The signaling events that result in NO-independent guanylate cyclase activation and thereby attenuation of the vasoconstrictor response to the stable thromboxane analogue U46619 are presently not known. In addition to the NAD(P)H oxidase-superoxide-H₂O₂ axis addressed in the preceding discussion, soluble guanylate cyclase activation via carbon monoxide generated by hemoxygenase(s) (30) might be operative. In recent studies in rat lungs, however, in which inhaled carbon monoxide was noted to interfere with HPV, its influence was noted to be independent of cGMP formation (33). Atrial natriuretic peptides are known to exert pulmonary vasodilation via stimulation of the particulate guanylate cyclase (34, 35). Further studies are clearly mandatory to clarify the nature of NO-independent guanylate cyclase activation counterbalancing the U46619-elicited pressor response and to resolve the question of why such activity is apparently not operative in limiting the hypoxia-induced pulmonary vasoconstriction.

In conclusion, the current data support an important role of NO-dependent guanylate cyclase activity in attenuating the vasoconstrictor response to alveolar hypoxia in rabbit lungs. In contrast, no evidence was obtained for a role of NO-independent cGMP formation in HPV, as suggested to occur in response to superoxide/H₂O₂ formation. In this feature the HPV differs from the vasoconstrictor response to thromboxane, which was noted to be attenuated via guanylate cyclase activity even under conditions of preblocked NO synthesis.